Optimized EGFR blockade strategies in EGFR addicted gastroesophageal adenocarcinomas

Purpose Gastric and gastroesophageal adenocarcinomas (GEA) represent the third leading cause of cancer mortality worldwide. Despite significant therapeutic improvement, the outcome of patients with advanced GEA is poor. Randomized clinical trials failed to show a significant survival benefit in molecularly unselected patients with advanced GEA treated with anti-EGFR agents. Experimental design We performed analyses on 4 cohorts: IRCC (570 patients), FMI (9397 patients), COG (214 patients) and INT (206 patients). Preclinical trials were conducted in patient-derived xenografts (PDXs). Results The analysis of different GEA patient cohorts suggests that EGFR amplification drives aggressive behaviour and poor prognosis. We also observed that EGFR inhibitors are active in patients with EGFR copy number gain and that co-amplification of other receptor tyrosine kinases or KRAS is associated with worse response. Pre-clinical trials performed on EGFR-amplified GEA PDX models revealed that the combination of an EGFR monoclonal antibody and an EGFR tyrosine kinase inhibitor was more effective than each monotherapy and resulted in a deeper and durable response. In a highly EGFR amplified non-responding PDX, where resistance to EGFR drugs was due to inactivation of the TSC2 tumor suppressor, co-treatment with the mTOR inhibitor everolimus restored sensitivity to EGFR inhibition. Conclusions This study underscores EGFR as a potential therapeutic target in gastric cancer and identifies the combination of an EGFR TKI and a monoclonal antibody as an effective therapeutic approach. Finally, it recognizes mTOR pathway activation as a novel mechanism of primary resistance that can be overcome by the combination of EGFR and mTOR inhibitors.